Trap/Cognitive Bias/No. 0688

Omission Bias

Omission bias is the tendency to judge harm from inaction less harshly than comparable harm from action without justification. Studied by Ilana Ritov and Jonathan Baron in vaccination decisions, it affects both choices and moral judgments and differs from status quo bias.

a trap: easy to walk into

Evidence: Well established
Read: 6 min
Links: 7 connections
Useful when: Deciding under uncertainty · Fairness and ethics · Group decisions and meetings · Judging people and hiring · Risk and safety

01You've seen this when…

in life
Your friend is about to miss a refund deadline. You know the date but say nothing; costing them the same money through a wrong tip would feel much worse.
at work
A manager keeps a hazardous shift schedule. Changing it feels like taking responsibility for what happens next; keeping it feels like staying out of the way.
out in the world
A flood barrier needs maintenance. Officials worry more about authorizing a repair that might fail than about the predictable damage if they authorize nothing.

02The idea

A bad outcome feels different depending on whether you brought it about by acting or allowed it by doing nothing. Giving someone harmful advice looks worse than withholding a warning. Approving a change that causes an injury feels worse than leaving a known hazard untouched.

Omission bias is giving harmful inaction an unjustified discount. It can affect both judgment and choice: you blame someone less for allowing harm, or avoid an action because causing harm would feel worse than allowing comparable harm.

The word unjustified matters. Actions and omissions can differ in consent, duties, intentions, knowledge and consequences. Those differences may support different judgments. The bias appears when doing nothing receives lighter treatment even after the relevant differences are accounted for.

Omission bias and status quo bias favor different things. Status quo bias favors the existing arrangement. Omission bias favors inaction, especially when intervention makes you feel responsible for a possible loss. The two often overlap; maintaining the existing arrangement and doing nothing can be different choices.

03Why it happens

Actions make responsibility easy to picture. When someone signs an order, makes a recommendation or presses a button, the harm has a visible starting point. An absent warning leaves less of a trail. That can lead people to underestimate inaction’s causal role.
Doing nothing supplies a convenient baseline. We use the world as it already stands as the baseline for judging an intervention. That comparison leaves other available futures out. Harm that occurs without intervention can start to look like background misfortune, obscuring the chance to prevent it.
Causing harm can feel harder to live with. People may anticipate more guilt or regret after a harmful action than after a harmful omission. Avoiding that feeling can take priority over reducing the harm itself.
Social and moral rules often distinguish doing from allowing. Many rules forbid directly harming people more strongly than they require preventing every harm. That distinction can have good reasons. The mistake is extending it beyond the cases those reasons support.

These explanations aren’t interchangeable. Someone may be protecting their reputation, following a moral principle, or misreading their causal role. The same inactive choice can come from different motives.

04A worked example

In 1991, Elisa Minsk and Jonathan Baron collaborated with Mark Spranca on a study using a hypothetical tennis matchup. One player knows that a food ingredient can make his opponent ill. In one version, he recommends the food. In another, he lets the opponent choose it without warning him. The scenarios hold key features, including harmful intent and knowledge, comparable.

What it looks like One player actively sabotages his opponent; the other merely fails to help. The recommendation gives observers a concrete act to condemn.

What’s actually going on Both players knowingly let their opponent face the harmful ingredient for competitive advantage. Participants nevertheless tended to judge the harmful action more harshly than the omission. The experimental scenario exposed a difference in judgment that couldn’t simply be explained by one player having innocent intentions.

What would have helped Explicitly comparing what each player knew and wanted, then assessing what each could have prevented before judging the recommendation versus the silence. That’s a reasoning prompt; establishing it as a cure would require evidence beyond this study. It also leaves room for a principled action–inaction distinction and calls for a defense of it.

05How to spot it

Only intervention needs a defense. You demand evidence before changing course but don’t examine the expected harm of continuing.
Responsibility disappears when nobody signs. A decision feels safer because there is no new approval, even though everyone knows what inaction permits.
Your imagined regret outweighs the outcome. You focus on how you would feel if your action went badly, not on how often each option goes badly.
The wording changes your verdict. Describing the same conduct as allowing harm rather than withholding protection noticeably softens your judgment.

06What to do instead

Put inaction on the options list. Give it the same treatment as every intervention: expected benefits, harms, uncertainty and who bears them. Leaving things alone is an option with consequences that deserve the same scrutiny.
Build the missing comparison. Use counterfactual reasoning: what is likely to happen if you act, and what is likely to happen if you don’t? Compare those futures, accounting for the harm that can occur without intervention.
Separate outcomes from moral constraints. First compare the consequences, using cost-benefit analysis where appropriate. Then examine consent, duties and rights. Better expected outcomes belong in the analysis, though they may be insufficient to justify an action.
Reverse the framing. Apply considering the opposite. Describe inaction as declining to prevent a foreseeable harm, then describe acting as accepting a risk to prevent it. Both descriptions need scrutiny; comparing them reveals what your original wording hid.
Name the reason for the distinction. If you still prefer inaction, identify the relevant difference. Lack of authority, uncertain effectiveness or a violation of consent can be reasons. The choice’s merits require evidence beyond feeling less implicated.

07When it isn’t omission bias

You aren’t responsible for preventing every harm you could conceivably prevent. Ability, knowledge, cost, authority and special duties all matter. A passerby and a lifeguard don’t have identical obligations. Declining an intervention you can’t perform safely may be sensible, not biased.

Actions can also cross boundaries specific to intervention. Treating someone without consent raises a separate problem from failing to provide treatment. Deontological ethics takes duties and constraints seriously; disagreement with a purely outcome-based judgment can be consistent with sound reasoning.

Reluctance to intervene can have explanations beyond omission bias, including when someone is deciding whether to vaccinate or make a repair. People may have different beliefs about the risks or distrust the information. Establish what they believe before explaining their choice.

The useful test is whether the judgment changes mainly because one outcome follows an action and the other follows inaction, after the relevant differences have been examined.

08Roots

In 1990, Ilana Ritov and Jonathan Baron put a parent’s nightmare into a decision study: a vaccine meant to protect a child could itself cause a fatal reaction. They asked how people weighed that treatment risk against the risk from disease.

Their paper, Reluctance to vaccinate: Omission bias and ambiguity, helped establish the modern psychological account. A death following vaccination could carry a special burden because the parent had authorized the intervention. Leaving the child unvaccinated could feel less blameworthy, even when it meant accepting greater danger. The study also examined uncertainty about risks, allowing for different explanations of reluctance.

The tennis scenarios followed in 1991, moving the question from protective decisions to moral judgment. Later work examined how reference points influence what counts as acting or omitting. Together, these studies made an old moral distinction into a testable psychological question: how much of our judgment tracks relevant differences, and how much tracks whether someone visibly did something?

09How solid is this?

Controlled studies document action–inaction differences in choices and moral judgments. Their size depends on framing, intentions, duties and uncertainty; the evidence does not establish that every preference for inaction is irrational.

10Connections

Often confused with Status Quo Bias, Default Effect
Countered by Counterfactual Reasoning, Cost-Benefit Analysis, Consider-the-Opposite Strategy
See alsoRegret Aversion, Deontological Ethics

11Origin and sources

Ilana Ritov and Jonathan Baron described omission bias in vaccination decisions in 1990. Mark Spranca, Elisa Minsk and Jonathan Baron extended its study to moral judgment in 1991.

[1]Ritov, I., & Baron, J. (1990). Reluctance to vaccinate: Omission bias and ambiguity. Journal of Behavioral Decision Making, 3(4), 263–277.
[2]Spranca, M., Minsk, E., & Baron, J. (1991). Omission and commission in judgment and choice. Journal of Experimental Social Psychology, 27(1), 76–105.
[3]Baron, J., & Ritov, I. (1994). Reference points and omission bias. Organizational Behavior and Human Decision Processes, 59(3), 475–498.

Suggest an edit· Updated 2026-10-02